Readme.txt for “Elite Rhetoric and the Running Tally of Party-Group Linkages”

Christoffer Hentzer Dausgaard & Frederik Hjorth.

CREATED: December 1, 2025

This folder includes all files necessary to produce the results of the above paper. The following details the content of each script and the order in which they should be run to reproduce all reported results in the manuscript and online Supporting Information. 


DATA FILES:

groupappeals.rds: original parliamentary speech dataset with annotated group appeals, at the mention level (Hjorth & Dausgaard 2025a)

press-appeals.rds: original press release dataset with annotated group appeals, at the mention level (Hjorth & Dausgaard 2025b)

V-Dem-CPD-Party-V2.rds: raw V-Party data on party positions (Lindberg et al. 2022)

BESIP.rds: raw survey data from British Election Study Internet Panel waves 1-25 (Fieldhouse et al. 2023)

BES1997.dta, BES2001.dta, BES2005.dta, BES2010.dta, BES2015.dta, BES2017.dta, BES2019.dta: raw survey data from British Election Study cross-section election surveys from 1997-2019 (Fieldhouse et al. 2022, Fieldhouse et al. 2019a, Fieldhouse et al. 2019b, Whiteley & Sanders 2014, Clarke et al. 2006, Clarke et al. 2003, Heath et al. 1999)

BESIP_clean.rds: cleaned version of BESIP.rds for analysis

BESrounds.rds: cleaned and merged version of BES survey files for analysis

fulldf.rds: dataset that merges groupappeals.rds, press-appeals.rds, V-Dem-CPD-Party-V2.rds, BESIP_clean.rds, and BESrounds.rds to combine date-stamped group appeals and survey datasets for analysis

CODE FILES:

prepare-surveydata.R: Cleans and harmonizes raw British Election Study (BES) survey data (BESIP.rds, BES1997.dta, BES2001.dta, BES2005.dta, BES2010.dta, BES2015.dta, BES2017.dta, BES2019.dta). Outputs two datasets, one for the BES cross-sections (BESrounds.rds) and one for the BES panel (BESIP_clean.rds)

merge.R: Merges survey data (BESIP_clean.rds and BESrounds.rds), speech data (groupappeals.rds) and party position data (V-Dem-CPD-Party-V2.rds) to construct individual-wave-dyad-level exposure variables that combine interview timing from surveys and dyad-mention level net valences into ‘exposure windows’ to return a full, merged datafile from 1997-2022 (fulldf.rds)

estimate.R: Reproduces all analyses, tables and figures in the main text and appendix using the full merged file (fulldf.rds) as well as the separate cross-section and panel files (BESrounds.rds and BESIP_clean.rds) and the annotated press releases file (press-appeals.rds)


OTHER FILES:

- codebook_besip.txt: Codebook for all variables in BESIP_clean.rds
- codebook_besrounds.txt: Codebook for all variables in BESrounds.rds
- codebook_groupappeals.txt: Codebook for all variables in groupappeals.rds
- codebook_press.txt: Codebook for all variables in press-appeals.rds
- codebook_fulldf.txt: Codebook for all variables in fulldf.rds
- customized_group_dictionary.xlsx: our dictionary for group mentions in speech
- Readme.txt: This file. 

Order to run code files for a full replication of results: 
 1. prepare-surveydata.R
	inputs: BESIP.rds, BES1997.dta, BES2001.dta, BES2005.dta, BES2010.dta, BES2015.dta, BES2017.dta, BES2019.dta
	outputs: BESrounds.rds, BESIP_clean.rds
 2. merge.R
	inputs: BESIP_clean.rds, BESrounds.rds, groupappeals.rds, V-Dem-CPD-Party-V2.rds
	outputs: fulldf.rds
 3. estimate.R
 	inputs: fulldf.rds, BESrounds.rds, BESIP_clean.rds, press-appeals.rds

Note: all results can be replicated using only "estimate.R" as the cleaned datasets produced by the other code files are included in the replication package. Files and data for constructing the original datasets groupappeals.rds and press-appeals.rds can be found on https://github.com/chdausgaard/running-tally-replication. 

Total time to run all code (steps 1-3) in author's computing environment: 631 minutes
- prepare-surveydata.R: 0.5 minutes
- merge.R: 623 minutes
- estimate.R: 7 minutes

Software: 
	R version 4.4.1 (2024-06-14)
	Computer Operating System: macOS Sonoma 14.1
	Computer Processor: Apple M1 Pro (8-core CPU: 6 performance cores, 2 efficiency cores)
	Computer Memory (RAM): 16.0 GB unified memory; 64-bit operating system, ARM-based processor

Software packages and version numbers:
    MASS_7.3.65
    tidyverse_2.0.0
    cowplot_1.1.3
    fixest_0.12.1
    marginaleffects_0.22.0
    modelsummary_2.2.0
    ggbrace_0.1.1
    lemon_0.5.0
    hexbin_1.28.4
    conflicted_1.2.0
    labelled_2.13.0
    data.table_1.16.2
    haven_2.5.4
    Hmisc_5.1.3

References for source data:

Hjorth, F., and C. H. Dausgaard. 2025a. "hoc-appeals (Revision 6f6f6c1)." Hugging Face. https://huggingface.co/datasets/groupappealslab/hoc-appeals. doi:10.57967/hf/7286.

Hjorth, F., and C. H. Dausgaard. 2025b. "press-appeals (Revision 4d80260)." Hugging Face. https://huggingface.co/datasets/groupappealslab/press-appeals. doi:10.57967/hf/7287.

Fieldhouse, E., J. Green, G. Evans, J. Mellon, C. Prosser, J. Bailey, R. de Geus, H. Schmitt and C. van der Eijk. 2023. “British Election Study Internet Panel Waves 1-25.”

Fieldhouse, E., J. Green, G. Evans, J. Mellon, C. Prosser, R. de Geus and J. Bailey. 2022. “British Election Study, 2019: Post-Election Random Probability Survey.”

Fieldhouse, E., J. Green, G. Evans, J. Mellon, C. Prosser, H. Schmitt and C. van der Eijk. 2019a. “British Election Study, 2017: Face-to-Face Post-Election Survey.” [data collection]. UK Data Service. SN: 8418, DOI: 10.5255/UKDA-SN-8418-1.

Fieldhouse, E., J. Green, G. Evans, J. Mellon, C. Prosser, H. Schmitt and C. van der Eijk. 2019b. “British Election Study, 2015: Face-to-Face Post-Election Survey.” [data collection]. UK Data Service. SN: 7972, DOI: 10.5255/UKDA-SN-7972-1.

Whiteley, P. F. and D. Sanders. 2014. “British Election Study, 2010: Face-to-Face Survey.” [computer file]. Colchester, Essex: UK Data Archive [distributor].

Clarke, H. et al. 2006. “British Election Study, 2005: Face-to-Face Survey.” [computer file]. Colchester, Essex: UK Data Archive [distributor].

Clarke, H. et al. 2003. “British General Election Study, 2001; Cross-Section Survey.” [computer file]. Colchester, Essex: UK Data Archive [distributor].

Heath, A. et al. 1999. “British General Election Study, 1997; Cross-Section Survey.” [computer file]. 2nd Edition. Colchester, Essex: UK Data Archive [distributor].

Lindberg, Staffan I., Nils Düpont, Masaaki Higashijima, Yaman Berker Kavasoglu, Kyle L. Marquardt, Michael Bernhard, Holger Döring, Allen Hicken, Melis Laebens, Juraj Medzihorsky, Anja Neundorf, Ora John Reuter, Saskia Ruth-Lovell, Keith R. Weghorst, Nina Wiesehomeier, Joseph Wright, Nazifa Alizada, Paul Bederke, Lisa Gastaldi, Sandra Grahn, Garry Hindle, Nina Ilchenko, Johannes von Römer, Steven Wilson, Daniel Pemstein and Brigitte Seim. 2022. “Varieties of Party Identity and Organization (V–Party) Dataset V2.”. URL: https://doi.org/10.23696/vpartydsv2


